top of page
cbdde9faadf0c49f07a366c812cc5882.JPG

Independent Study

Ergonomics of chairs: Utilizing pressure distribution data to modify seat and backrest

By. HU Xinhan

This project will build a low-cost, self-made pressure distribution sensor system and modify the seat and backrest of the wheelchair to support an appropriate sitting posture in a sedentary situation.

prolonged sitting.jpg

Background

Prolonged sitting, especially with poor posture, can lead to health issues such as scoliosis, muscle atrophy, and pressure sores. Wheelchair users are particularly vulnerable to these problems, significantly impacting their quality of life. Currently, there is no systematic medical cure, and field surveys show that patients often rely on traditional Chinese therapies or makeshift solutions (e.g., ropes and fabrics) to stabilize themselves in wheelchairs.

A well-designed wheelchair—with proper sizing, supportive cushions, backrests, and armrest positioning—is essential. However, the market offers limited ergonomic guidance, leaving users to spend excessive time and effort selecting or modifying wheelchairs. Custom wheelchairs are often prohibitively expensive for most individuals.

While ergonomic office chairs are widely available, advanced pressure-mapping technology (e.g., pressure mats) remains costly, data-intensive, and inaccessible to standard wheelchair manufacturers and clinics.

Inspiration / Reference

TekscanConforMat system.jpg
office chair.png

Equipped with 2048 sensors, the commercially available Tekscan ConforMat system is a leading pressure-mapping solution. Prior research (Bilge et al., 2007; Martinaitis et al., 2018) has simplified such systems into affordable, self-made sensor arrays, proving feasibility for posture monitoring in office chairs.

Purpose

  • Develop a low-cost, self-made pressure distribution sensor system for wheelchairs.

  • Analyze the relationship between pressure data and sitting posture.

Methodology

Control variables:

  • Sitting position (front/middle/back of cushion)

  • Pelvic rotation (forward/straight/backward)

  • CG position in width (left/straight/right)

 

Data analysis and visualization:

  • Arduino for pressure data collection

  • MATLAB for data visualization

Literature Review​

pressure map.png
grad_function.png
pelvic.jpg

  • Existing research uses detailed pressure data to train posture-prediction algorithms, but this requires extensive datasets (limited by time/budget).

  • Most used static pressure distribution parameters: Peak pressure (PeakP), mean pressure (MeanP), standard deviation of the pressure distribution (STDP), total contact area (Area), force (Force), maximum gradient (MaxGrad), mean gradient (MeanGrad) and the standard deviation of the gradient (STDGrad)
    *The gradient (2) was defined by the geometrical addition of the pressure derivative of the two sensor mat directions (x, y) and resulted in a 15 by 15 matrix

  • The relationship between the center of gravity and the sitting pressure: In the anterior position the center of gravity is in front of the ischial tuberosities, and the feet transmit more than 25% of the body weight to the floor. In the posterior position, the center of gravity is above or behind the ischial tuberosities, and less than 25% of the body weight is transmitted by the feet.

Sensor System

​Components:​

  • Flexible thin-film pressure sensors (5kg–50kg) ×5 (includes single-channel module + IMS_C10A_50kg sensor)

  • Supporting materials: Wires, breadboard, silicone gasket, cardboard

  • Coverage area: 5×5 matrix (20cm × 15cm)

10af2d641d1c350b1d616a379ef93f9e.JPG

Challenges & Solutions

Issue:

  • Small sensor size made it difficult to capture accurate data when directly sitting on them. And the data variation is small. Sometimes, the occasional signal loss (outputting 0) happened.

 

Solution:

  • Tested low-cost cover materials to amplify signals:

Result:

  • The dynamic pressure distribution graphs of different sitting postures:

Issue:

  • Small sensor size made it difficult to capture accurate data when directly sitting on them. And the data variation is small. Sometimes, the occasional signal loss (outputting 0) happened.

 

Solution:

  • Tested low-cost cover materials to amplify signals:

Result

The dynamic pressure distribution graphs of different sitting postures:

Sit straight in the middle

Lean backward

Lean forward

Lean left

Lean right

The result enables personalized seat adjustments to support healthy posture and enhance comfort.

Significance

This research benefits wheelchair users by:

  • Providing a low-cost system to guide posture adjustments, reducing spinal strain and muscle injuries.

  • Facilitating data sharing with physicians for better rehabilitation insights.
     

For our project, it enhances wheelchair comfort and usability while enabling affordable customization—supporting modular design and user-platform development.

ISDN2001/2002: Second Year Design Project

bottom of page